AI tool claims 97% efficacy in preventing ‘address poisoning’ attacks

Cryptocurrency address poisoning, a costly scam where attackers use similar-looking wallet addresses to trick users into sending funds, is now being combated by an AI-powered detection system. Developed by cybersecurity firm Trugard and onchain trust protocol Webacy, this tool leverages a supervised machine learning model trained on live transaction data. The system boasts a 97% success rate in identifying known attack cases, significantly improving upon existing methods.

Address poisoning exploits the user’s tendency to rely on partial address matching, leading to accidental transfers to attacker-controlled wallets. A study revealed over 270 million poisoning attempts on BNB Chain and Ethereum between July 1, 2022, and June 30, 2024, resulting in $83 million in losses from 6,000 successful attacks.

The new AI system addresses the limitations of traditional Web3 attack detection, which often rely on static rules and basic filtering that struggle to keep pace with evolving attacker tactics. Trugard and Webacy’s approach utilizes sophisticated algorithmic feature engineering and onchain analytics, combined with behavioral context analysis. This allows the system to learn and adapt to new attack patterns, providing a more robust and dynamic defense.

A key differentiator is the system’s emphasis on context and pattern recognition. Webacy co-founder Maika Isogawa highlights AI’s ability to detect patterns beyond human capabilities. The model was trained using both real-world data and synthetically generated data simulating various attack scenarios, enabling the system to generalize and remain effective against new, unseen attacks.

The synthetic data generation layer allows for continuous testing and improvement of the model, ensuring its ongoing robustness and adaptability. This proactive approach ensures the system stays ahead of evolving attack methods, offering a significant advancement in cryptocurrency security. The collaboration between Trugard and Webacy represents a notable step towards mitigating the substantial financial losses associated with address poisoning attacks.